UNIX for Beginners Questions & Answers

I have some data that looks like this:

Code:

PXD= ZW< 1,6
QR> QRJ== 1,2(5)
QR> QRJ== 4,1(2)
QR> QRJ== 4,2
QRB= QRB 4,2
QWM QWM 6,2
R<C ZW< 11,2
R<H= R<J= 6,1
R>H XZJ= 1,2(2)
R>H XZJ= 2,6(2)
R>H XZJ= 4,1(2)
R>H XZJ= 6,2
RDP RDP 1,2

What I would like to do is if fields $1 and $2 are identical in consecutive records, I would like to print the values of field $3 in those lines, separated by semi-colons. Thus the desired output would look like this:

Code:

PXD= ZW< 1,6
QR> QRJ== 1,2(5); 4,1(2); 4,2
QRB= QRB 4,2
QWM QWM 6,2
R<C ZW< 11,2
R<H= R<J= 6,1
R>H XZJ= 1,2(2); 2,6(2); 4,1(2); 6,2
RDP RDP 1,2

My desire would be to accomplish this with awk, but am open to have this work with other rudimentary unix/linux programs (e.g., uniq, paste, tr, et al).

I can find duplicates easy enough, it is the writing to desired output that is stumping me. I have tried numerous attempts with post-incremented arrays, but nothing I'm doing seems to work and I'm so new to this that I'm hoping for an example from which I can inductively learning from.

Thank you so much in advance.

Hi, there are many examples on these forums. Here is one way you could try:

Code:

awk '{n=$1 OFS $2} n!=p{if(p) print p, r; r=$3; p=n; next}{r=r "; " $3} END{if(p) print p,r}' file

Scrutinizer, this is just some brilliant stuff here! Thank you so much.

If I can impose upon you a bit more would you be willing to either point me in the direction of some of the other examples you mention or even offer a few bullet points of explanation of this code you've offered? In particular, I'm a bit confused on the were you define n!=p when p does not seem to have been defined yet (this goes for r as well).

Hello jvoot,

Could you please try following and let me know if this helps you.

Code:

awk 'prev && prev!=($1$2){print val;val=""} {val=val?val OFS $3:$0;prev=$1$2} END{print val}' OFS=";"   Input_file

Thanks,
R. Singh

As a starting point, look at the threads at the end of this page in the box with the heading "More UNIX and Linux Forum Topics You Might Find Helpful".

In awk, if a variable is used before any value is assigned to it, the value of that variable will be an empty string if it is used as a string or it will be 0 if it is used as a number.

Thanks Don. Unfortunately I noticed those suggested threads after I had responded to Scrutinizer. I really appreciate your own explanation here too as I continue to approach awk from a discipline far removed from computer science.